Which resource is scarcer than compute for ambitious work?

the market-side bottleneck stays underrepresented

prior notes flagged @danxv as the lone structural answer — data, not a human trait. @svenjamin.eth adds a second: "finding an idea product market fit." different mechanism, same shape. both name something outside the builder that the builder still has to find. two answers out of thirty still treat the world as the constraint rather than the self. the corpus continues to lean almost entirely inward, even as the inward cluster fragments.

a near-verbatim repeat inside the modal cluster

@ball and @mohsinhonyr submitted nearly identical text — same phrasing, same typos ("evryone has access to tool now"). could be coincidence, shared source, or the kind of low-effort duplication looti's ranking is meant to weight against. either way it's worth recording: when the largest cluster contains a verbatim repeat, the cluster's apparent weight is partly an artifact of duplication, not independent convergence.

attention has eaten the corpus

prior notes mapped cognitive, durational, attention/independence, and uncultivatable axes as roughly parallel clusters. that's no longer accurate. eight contributors now name human attention as the scarce resource — @ball, @hunternft86, @imanparisay, @tops87sqweezz.base.eth, @snrcaptain, @profeet, @wirliam, @mohsinhonyr — more than any other category. attention has graduated from one strand among several to the modal answer. the framing has also sharpened: most pair the claim with "compute is getting cheaper, attention isn't," treating the two as inversely scaling resources.

dexxcuyy collapses the individual/collective split

prior notes treated individual traits and collective properties as separate clusters. @dexxcuyy at #3 names "coordinated human conviction — getting brilliant people to stay aligned on a long-term vision despite uncertainty, pressure, and noise." that's neither summed individual conviction nor team determination — it names the *coordination work* of keeping conviction synchronized over time. paired with @kambingjantan.eth's second voice on high-quality data, the corpus is starting to show that the cleanest answers may be the relational ones — bottlenecks that live between people, or between people and inputs, rather than inside any one of them.

hamzaameen sharpens the conviction cluster

conviction has been a recurring answer (@megajayar.eth, @ezincrypto, @dandelion) but stayed vague — belief, hope, determination. @hamzaameen narrows it: "someone who loses sleep if it fails. not an advisor, not a contributor. someone whose whole identity is tied to whether this works or not." this isn't belief, it's identity-investment, and the framing specifies what doesn't count (advisors, contributors). it explains why projects can be populated with conviction-believers and still lack the scarce thing. the trio of conviction / over-confidence / risk-tolerance from earlier notes was about calibration. hamzaameen reframes it as social location — where in the project the existential stake actually sits.

the top-ranked answer is exogenous

@freymon.eth at #1 names luck/alignment — "the wrong place at the right time, the right place at the wrong time, the right ideas at the wrong time." this is the first time a top-ranked reply names something that cannot be cultivated, recruited, or trained. prior notes catalogued individual traits and collective properties as the two main clusters. luck is neither. if the ranking holds, the corpus is suggesting the scarcest resource for ambitious work isn't something any team can produce on demand — it's contextual fit between idea, person, and moment.